From Data To Diagnosis
Breast cancer remains the most common tumor disease among women in the European Union.
women in the EU are diagnosed with breast cancer each year - and the number is rising...
1 in 11 women in the EU 27 will develop breast cancer before the age of 74.
source: europa-donna.org
Despite the availability of screening programs a significant number of breast cancer incidents is not discovered - or not discovered early enough. Early detection is a key factor in reducing mortality and increasing the chances of a successful recovery.
However there are some barriers that hinder effective screening. For example, among immigrant women there is a lower participation rate in mammography programs due to language barriers, cultural misunderstandings and a lack of awareness about the importance of early detection.
Another challenge arises for women with smaller breasts, as mammograms may struggle to capture sufficient tissue for accurate results. AI technology has the potential to improve image quality, enabling the detection of subtle abnormalities and enhancing the reliability of early breast cancer diagnosis.
Mag. Dr. Maria Bernathova is a radiologist specializing in breast health. In 2016 she opened her own practice in Vienna, where she utilizes AI in her work.
Dr. Liz O'Riordan on AI EDITH
The low mammography rate is not only a problem in Austria. In the United Kingdom, too, the government is trying to increase the participation rate. Therefore a large-scale implementation of AI for mammogram analysis is planned. AI EDITH will enable faster and more accurate detection of abnormalities, increasing the effectiveness of early breast cancer diagnosis.
AI-Solutions: How Is AI Used?
Doctors have used computers to help them detect breast cancer for years. But many cases still go undetected.
The synergy between AI and radiology holds immense potential for advancing breast cancer care and improving outcomes for patients. AI offers the promise of earlier and more accurate detection, personalised risk assessment, streamlined workflows, and improved access to care, particularly in underserved communities.
For example, AI can analyse mammograms with greater sensitivity than the human eye, potentially detecting subtle changes that indicate early-stage cancer. It can also integrate data from various sources, such as mammograms, patient history, and genetic information, to create a more comprehensive risk profile for each woman, enabling tailored screening strategies.
Due to AI, doctors can work more efficiently. AI can provide quantitative analyses that are not subject to human bias, making data-driven calls for unclear mammograms that could be interpreted differently. AI-powered software can automate the interpretation of breast mammograms, ultrasounds, and MRI-scans to get patients their results faster.
Mammogram of a ductal carcinoma
Furthermore, AI-powered imaging may also improve breast cancer care in low-resourced or rural areas, where women often face significant challenges, including a shortage of radiologists and long travel distances to specialised centres. AI-systems enable remote interpretation of imaging, facilitating timely diagnosis and treatment for patients regardless of where they live.
On this map of Austria the differences between rural and urban areas can be seen. Every blue dot marks a mammography-institute.
source: www.frueh-erkennen.at
But how does AI work?
To train AI to detect breast cancer, technicians have to input data from hundred of thousands of mammograms. Then the AI software engineers create a mathematical representation of what a unobtrusive mammogram should look like and what a mammogram with cancer looks like.
The AI system then checks each image against the standards to distinguish between healthy and unhealty breasttissue.
At last, AI can analyse patient data, including family history, genetic markers, lifestyle factors, and mammographic density to identify women at high risk, who may benefit from more frequent or specialised screening. This personalised risk assessment can lead to earlier detection and more tailored prevention strategies.
While this is very promising, it also raises concers. So what are the limits and are there any regulations?
Dipl. Ing. Philipp Seeböck specializes in Articifial Intelligence and Biomedical Research. He researches on the topic of individualized breast cancer MRI screening in high-risk women enabled by machine learning.
National Screening Program
Since 2014 Austria has a national program for the early-detection of breast cancer, frueh-erkennen.at, meaning "to recognise early". With a low threshold for getting access to the program, women in Austria between 45 and 74 are invited to have an early-detection mammogram every two years.
Let us have a look at current data from Austria.
source: www.frueh-erkennen.at
In addition to the mammogram, an ultrasound examination is included in the screening program. Participation in the early-detection program is not mandatory. A letter regarding the examination-appointment is automatically sent to the women by the government.
The Future Of AI In Breast Screening
Looking to the future, the integration of artificial intelligence in breast cancer detection holds great promise for improving screening accuracy and accessibility across Europe. As AI technologies continue to advance, they are expected to refine diagnostic processes, reduce human error, and offer more personalized treatment options. With ongoing research and wider adoption, AI could play a pivotal role in bridging gaps in healthcare, ensuring that early detection and timely interventions become more efficient, particularly in rural and underserved areas. The future of breast cancer care in Europe is poised for significant transformation, driven by innovation and enhanced by AI’s potential.
However, realising this potential requires addressing significant challenges. More rigorous clinical trials are needed to ensure that AI systems are safe and reliable for use as a second or even third reader of mammograms, in addition to a radiologist.
DDr. Friedrich Semturs is a medical expert at the Medical University of Vienna, specializing in biomedical engineering. He focuses extensively on artificial intelligence in breast diagnostics.
AI can only be as good as the information that is used to train the technology. Therefore, data bias is a major concern. If AI algorithms are trained on data that doesn't represent the entire population, they may not perform accurately for all patients, potentially exacerbating existing health disparities.
I am totally convinced, that the ones who will benefit the most from Artificial Intelligence screenings, will be the women.
Maria Bernathova
Furthermore, the cost of implementing AI systems can be substantial, requiring significant investment in software, hardware, and training for healthcare professionals. Ethical considerations must also be addressed, including questions of liability if an AI system makes an error, ensuring patient privacy, and maintaining human oversight in the diagnostic process.
Despite these challenges, the potential of AI to revolutionise breast cancer screening is undeniable. Continued research, collaboration between clinicians and AI developers, and careful attention to ethical and practical considerations will be essential to realise this potential and improve the lives of women across Europe.
In conclusion, breast cancer remains a critical health concern for women, necessitating effective screening, early detection, and timely treatment. While national programs like Austria’s have improved access to early detection services, disparities in participation and diagnostic accuracy persist.
The integration of artificial intelligence into breast cancer screening has the potential to address these challenges by enhancing image analysis, expediting diagnoses, and improving care in underserved areas. As AI technologies continue to evolve, they hold promise for reducing breast cancer mortality and improving outcomes for women across Europe.
source: Lily Williams https://www.flickr.com/photos/diethylstilbestrol/22406036396
CHECK YOURSELF
All genders can be affected by breast cancer. Know your body and talk to your healthcare provider about any changes you notice.
Regular self-exams are a valuable tool for early detection, but they are not a substitute for professional medical advice or regular screenings, including mammograms.
If you are a trans man, remember that breast tissue may still be present even after top surgery, and you should follow your healthcare provider's recommendations for screenings and self-exams.
Remember to consult your healthcare provider with any concerns or questions.
Regards
This multimedia story was created within one week in Vienna, during the International School of Multimedia Journalism – a joint program for journalism students from Austria, Denmark, Georgia and Ukraine.
Made by:
Anja Bauer (Austria)
Sevgi Ismayilbayli (Georgia)
David Lemser (Denmark)
Anastasiia Ordynets (Ukraine)
Dominik Volb (Austria)
Mentor:
Saskia Hödl (Austria)
Special thanks to:
Dr. Maria Bernathova
Philipp Seeböck
DDr. Friedrich Semturs
frueh-erkennen.at
europa-donna.org